ESTRO 2024 - Abstract Book

S4329

Physics - Intra-fraction motion management and real-time adaptive radiotherapy

ESTRO 2024

Purpose/Objective:

Expiration gating is used as respiratory motion management strategy for pancreatic cancer. However, expiration gating suffers from low time efficiency 1 , resulting in long treatment fraction durations. Non-invasive mechanical ventilation (NIMV) can be used to mechanically alter the patients’ breathing pattern 2 to lengthen the expiration phase during expiration gating radiotherapy resulting in increased time efficiency. The purpose of this study was to quantify the effect of NIMV on time efficiency and required margins for expiration gating in pancreatic cancer.

Material/Methods:

After ethics committee approval and informed consent, three healthy volunteers enrolled. Subjects practiced being non-invasively ventilated at their free breathing frequency (~10-12 breaths per minute) and at 5 breaths per minute (brpm) with prolonged expiration. In two subsequent MRI sessions, we acquired MRI images during free breathing (FB) and mechanical ventilation at their typical own free breathing frequency, and at 5 brpm. We acquired single–slice sagittal images, intersecting the pancreas head and the abdominal wall, at 2 Hz for six minutes. Using deformable image registration we acquired deformation vector fields, showing motion of every pixel from a reference image to each other image. On the reference image, we delineated the pancreas and abdominal wall and quantified motion of the pancreas in cranio-caudal direction and of the abdominal wall in anterior-posterior direction over the entire six minute acquisition. On the abdominal wall motion, we determined a base level for the gating window by selecting the mean position of the first five stable end-expiratory positions of the abdominal wall. Next, we constructed a gating window around this mean position in accordance with our clinical protocol. The anterior gating window bound was set 3 mm anterior when abdominal wall motion was larger than 10 mm, and 2 mm anterior when abdominal wall motion was between 10 and 8 mm (denoted as +3 or +2). The posterior gating window bound was set at 1mm if no variation was observed in the first-five end-expiration positions, and at 2mm if some variation was observed in the first five end-expirations (denoted as -1 or -2). Therefore, the gating window size was 4 or 5mm. We determined the time efficiency as the percentage of abdominal wall positions within the gating window. Furthermore, we assessed the variation in pancreas position when the abdominal wall was within the gating window as an estimation for required margins in cranio-caudal direction.